An Overview of Core CI Technologies

CITS7212: Computational IntelligenceAn Overview of Core CI Technologies

An Overview of Core CI Technologies

Lyndon While


Evolutionary Algorithms (EAs)

Reading:

X. Yao, “Evolutionary computation: a gentle introduction”, Evolutionary Optimization, 2002

T. Bäck, U. Hammel, and H.-P. Schwefel, “Evolutionary computation: comments on the history and current state”, IEEE Transactions on Evolutionary Computation 1(1), 1997

CITS7212: Computational Intelligence

The Nature-Inspired Metaphor

Inspired by Darwinian natural selection:– population of individuals exists in some environment– competition for limited resources creates selection

pressure: the fitter individuals that are better adapted to the environment are rewarded more than the less fit

– fitter individuals act as parents for the next generation– offspring inherit properties of their parents via genetic

inheritance, typically sexually through crossover with some (small) differences (mutation)

– over time, natural selection drives an increase in fitness

EAs are population-based “generate-and-test” stochastic search algorithms

An Overview of Core CI Technologies 3


Terminology

Gene: the basic heredity unit in the representation of individuals

Genotype: the entire genetic makeup of an individual – the set of genes it possess

Phenotype: the physical manifestation of the genotype in the environment

Fitness: evaluation of the phenotype at solving the problem of interest

Evolution occurs in genotype space based on fitness performance in phenotype space


5


The Evolutionary Cycle


Termination

Initialisation

Population

Parents

Offspring

Parent selection

Survivor selection

Genetic variation(mutation and recombination)


245 6 8 5

XX

An Example



EA Variants

There are many different EA variants/flavours:– differences are mainly cosmetic and often irrelevant – the similarities dominate the differences

Variations differ predominately on representation:– genetic algorithms (GAs): binary strings– evolution strategies (ESs): real-valued vectors– genetic programming (GP): expression trees– evolutionary programming (EP): finite state machines

and also on their historical origins/aims, and their emphasis on different variation operators and selection schemes



Designing an EA

Best approach for designing an EA:– choose a representation to suit the problem, ensuring

(all) interesting solutions can be represented– create a fitness function with a useful search gradient– choose variation operators to suit the representation,

being mindful of any search bias – choose selection operators to ensure efficiency while

avoiding premature convergence to local optima– tune parameters and operators to the specific problem

Need to balance exploration and exploitation:– variation operations create diversity and novelty– selection rewards quality and decreases diversity



Neural Networks (NNs)

Reading:

S. Russell and P. Norvig, Section 20.5, Artificial Intelligence: A Modern Approach, Prentice Hall, 2002

G. McNeil and D. Anderson, “Artificial Neural Networks Technology”, The Data & Analysis Center for Software Technical Report, 1992


The Nature-Inspired Metaphor

Inspired by the brain:– neurons are structurally

simple cells that aggregate and disseminate electrical signals

– computational power and intelligence emerges from the vast interconnected network of neurons

NNs act as function approximators or pattern recognisers which learn from observed data


Diagrams taken from a report on neural networks by C. Stergiou and D. Siganos


The Neuron Model

A neuron combines values via its input and activation functions

The bias determines the threshold needed for a “positive” response

Single-layer neural networks (perceptrons) can represent only linearly-separable functions


n

j

jiji aWga0

,

g

ia

iW ,0

ijW ,ja

10 a

InputLinks

Output Links

Output

Bias Weight

ActivationFunction

InputFunction


Multi-Layered Neural Networks

A network is formed by the connections (links) of many nodes – inputs to outputs through one or more hidden layers

Link weights control the behaviour of the function represented by the NN– adjusting the weights

changes the encoded function



Multi-Layered Neural Networks

Hidden layers increase the “power” of the NN at the cost of extra complexity and training time:– perceptrons capture only linearly-separable functions– an NN with a single (sufficiently large) hidden layer can

represent any continuous function with arbitrary accuracy– two hidden layers are needed to represent

discontinuous functions

There are two main types of multi-layered NNs:1. feed-forward: simple acyclic structure – the stateless

encoding allows functions of just its current input

2. recurrent: cyclic feedback loops are allowed – the stateful encoding supports short-term memory



Training Neural Networks

Training means adjusting link weights to minimise some measure of error (the cost function)– i.e. learning is an optimisation search in weight space

Any search algorithm can be used, most commonly gradient descent (back propagation)

Common learning paradigms:1. supervised learning: training is by comparison with

known input/output examples (a training set)

2. unsupervised learning: no a priori training set is provided; the system discovers patterns in the input

3. reinforcement learning: training uses environmental feedback to assess the quality of actions



Questions?

Documents

An Overview of Core CI Technologies